| 1 |
I'm just getting up to speed on the 2.6.25 kernel. I've got it running, |
| 2 |
thanks to some hacking from VMware gurus (see Bugzilla |
| 3 |
https://bugs.gentoo.org/show_bug.cgi?id=218250 for the details.) I'm |
| 4 |
very interested in some of the new performance monitoring features. |
| 5 |
|
| 6 |
In particular, |
| 7 |
|
| 8 |
1. I've "discovered" that the block device partition statistics have |
| 9 |
been enhanced, and that there may be some kernel configuration settings |
| 10 |
that need to be documented. This is what I'm personally working on at |
| 11 |
the moment; I'm writing a paper for Computer Measurement Group on the |
| 12 |
block layer statistics, so I will probably hunt this one down myself. |
| 13 |
|
| 14 |
2. There is a tool called "latencytop" at http://www.latencytop.org/. It |
| 15 |
looks like the kernel patches to support it are in 2.6.25. How does one |
| 16 |
go about requesting "latencytop" in the Portage tree, dependent, of |
| 17 |
course, on 2.6.25 being stabilized? |
| 18 |
|
| 19 |
3. The process memory map statistics have been improved. I probably |
| 20 |
won't even look at this one, but if someone else is interested in |
| 21 |
"pioneering", there are some userspace tools at |
| 22 |
http://selenic.com/repo/pagemap/. |
| 23 |
-- |
| 24 |
gentoo-kernel@l.g.o mailing list |
Archives